Prefixing commands solves the namespace problem and discoverability (at least partly). I use a slightly more sophisticated method, which helps me remember which custom utilities are available and how to use them: sd [1], a light wrapper written for zsh that, in addition to namespaces, provides autocompletion, custom help texts + some other QoL enhancements. Can definitely recommend if you're looking for something a bit more fancy.
The Git plugin is great for single-device backup IMO, but not great for device sync or collaboration.
I've been working on making Obsidian "work for work" with a real-time collaboration plugin called Relay [0]. We use CRDTs for conflict resolution between users/clients and it also happens to remove a ton of headaches for device-to-device sync as well.
Our collaboration server can be run on-premise and we also just open sourced a Git Sync connector so you can do google-docs style collab via Obsidian+Relay but still have the merged documents end up in git (and plug into (Markdown + git)-centric publishing workflows like Mintlify and Quartz.
The whole Obsidian ecosystem feels really electric right now.
This is fantastic work. The focus on a local, sandboxed execution layer is a huge piece of the puzzle for a private AI workspace. The `coderunner` tool looks incredibly useful.
A complementary challenge is the knowledge layer: making the AI aware of your personal data (emails, notes, files) via RAG. As soon as you try this on a large scale, storage becomes a massive bottleneck. A vector database for years of emails can easily exceed 50GB.
(Full disclosure: I'm part of the team at Berkeley that tackled this). We built LEANN, a vector index that cuts storage by ~97% by not storing the embeddings at all. It makes indexing your entire digital life locally actually feasible.
Combining a local execution engine like this with a hyper-efficient knowledge index like LEANN feels like the real path to a true "local Jarvis."
Bill Dally from nvidia introduced a log representation that basically allows to replace a multiplication with an add, without loss of accuracy (in contract to proposal above)
[1] https://github.com/ianthehenry/sd